Subject: Re: "A combined analysis ..."
From: Ivan De Mitri (ivan.demitri@le.infn.it)
Date: Mon Feb 05 2001 - 06:28:02 EST
Ciao Marco,
here you can find the answers to your
comments and/or suggestions.
As far as your general comment is concerned...
If one is concerned with the ~2-3 candidates/yr
that are left to the CR-39 analysis, just apply
the "standard" cut in the dL/dx vs beta plane
and they are washed out (as indicated in the paper),
without the need for a two face analysis.
We should say that the general philosophy of the
analysis was to use all the subdetectors at the
same time with the lowest possible threshold in
each of them.
We decided to follow a conservative approach,
as a general characteristic of the analysis, at
the very beginning of this work (~7 yr ago).
These decisions also reflected the need for this kind of
analyses that was expressed by the rare particle wg,
and the spokesmen.
At the same time a similar work started in the U.S.
for the low beta region.
Work on other two combined analyses (one for the
whole beta range and one for the fast monopoles alone)
followed in subsequent years.
The analysis and its progress were presented
in all the collaboration meetings from '94,
a 40 pages memo was written in 97,and the results
were show in several conferences.
Up to now all the data were analyzed up to sep.99,
and we are going to analyze the last year.
Now we do not have the possibility to change the
analysis cuts and re-analyze the whole data sample.
List of Answers:
1) ok, done.
2) This is the number published in the paper
on fast monopole search with PHRASE and
ERP (R.Liu) . Astroparticle Physics 6 (1997) 112.
Is it wrong ?
3) Sorry, we don't understand your point.
4) ok, done.
5) ok, done.
6) ok, done.
7) ok, done.
8) Yes of course, this is already discussed in the text.
This sentence just presents the right part of the figure.
9) The same criteria apply to each of the fired counters.
10) ok, done.
11) As explained in the text this is the mean value of a gaussian
fit applied in each Lp slice. Since these behaviours have been
obtained by using large statistics, the mean and the truncated
mean just coincide. The concept of truncated mean is useful
(and used) on an event by event basis, where the statistic
is reduced to the number of fired streamer planes (with
a given cls).
12) ok, done.
13) Yes, the density of the information given in this part of the paper
is high. This is the result of our efforts trying to reduce
the paper length.
14) ok, done.
15) We don't see any contradiction in these statements.
16) The source data used for these figures are somewhere
in a backup tape.
We could put the arrows in the figure by just using
photoshop or some other graphics tool. By the way the
quality of the final picture might be worse
(ps <-> gif or jpg conversions).
Is it really needed ? As explained in the text, the cut
is defined to leave 90% of the monopoles on the right.
17) Here are the dE and dE/dX values of these events, as
measured by the ERP:
dE(MeV) de/dx (MeV/cm)
166 7.3
168 3.7
189 9.6
181 6.7
190 7.8
181 5.8
The threshold on the streamer T.o.F. is such that beta
is > 5*10^-3 at the level of event selection.
Even with this information alone, i.e. ERP dE/dX and streamer-beta
(> 5*10^-3), the events are vary far from the foreseen monopole
light yield curve.
If in some of these 6 events at least two ERP boxes were present
along the track, a better beta estimate can be obtained.
As a consequence of the previous sentence this is useless.
However this (and the event display) can be recovered by
re-analyzing the 6 events.
Let us know,
Ivan & Fausto
Marco Grassi wrote:
> Ciao Ivan,
>
> these are my comments on the article "A combined analysis technique
> for the search..."
>
> As you know, I have a general comment on the article philosophy, and
> I must repeat the argument here.
>
> In Macro we have two "active" monopole subdetector, the streamer
> tubes and the scintillator counters. These have been designed even to
> perform a stand-alone search for monopoles: the idea was to have a redundant
> threefold signature of a monopole event. The conclusion is that
> we can discard a candidate if just one subdetector has a null signature
> (probably in an analysis with more than one subdetector we could require
> more than one null signature ...) .
> In this paper we show a first attempt (may be also the last) to use
> both active subsystems, but the chosen approch is very conservative.
> The selection creteria, at least for the scintillators, require a
> modest energy release, if compared with the expectation, and a single
> counter information is sufficient to caracterize an event.
> The result is a number of good candidates that we must discard
> by using the CR39.
> I belive that e slightly higher energy cut ad at least 2 counters in
> the coincidence could bring the number of candidates to zero, without
> affecting the scintillators efficiency too much and without decreasing the
> geometrical acceptance because at lest 6 planes of streamer tubes are
> required.
> This comment is open for any kind of discussion.
>
> My editorial comments for the article are following.
>
> 1) page 1
> " ... an analysis technique using all of the Macro detector's
> capabilities"
> Well, not "all" capabilities have been used, I suggest:
> " ... an analysis technique using all of the Macro sub-detectors"
>
> 2) page 3
> " For the MACRO scintillator ... = (11.60 +/- 0.06) * 10**-3 ..."
> Just for my ignorance: who did measure the saturation constants with
> this accuracy? I remember R. Liu. Is it his number?
>
> 3) page 3
> " These discrete mechanisms produce showering events ... A single
> subdetector analysis would suffer ... a series of very stringent citeria"
> These criteria could be restricted a little bit more!
>
> 4) page 4
> "..distribution produced by a carefully selected set of cosmic muon
> events."
> We should define the set and then justify the choice. The adlective
> "carefully" could be avoided.
>
> 5) page 16 (fig1 caption)
> ".. the position of the hit along the scintillator .."
> substitute with
> ".. the position of the particle crossing along the scintillator .."
>
> 6) page 5
> " Near the scintillator ends this procedure fails by underestimating the
> energy lost by a particle."
> I agree. (The "standard" attenuation curve, used in all analysis,
> badly fails near the PMTs. The supposed 1/r**2 dependence is partially
> attenauted by the fact that one of the 2 PMTs when the event is very
> close to the counter end doesn't see the light!)
>
> 7) page 16 (fig2 caption)
> "All the effects have been taken into account"
> Remove this statment.
>
> 8) page 16 (fig2 caption)
> "... the cut applied to exclude corner clipping tracks and
> border effects ..."
> The reason for the 20 cm cut is already discussed in the text
> (see point 6)).
>
> 9) page 5
> "The pathlength of the particle in the ... greter than 10 cm".
> How do you treat the crossing of two adjacent countes?
>
> 10) page 7
> "As expected, the QTP charge distribution ... order of 20-30%"
> The definition of truncated mean should be enphasized and the entire
> period could be rephrased.
>
> 11) page 7
> "The dependence of the charge on Lp comes out to be: ..."
> This charge is the "truncated mean"? If not, the entire paragraph
> is confusing: at the end you use the Gamma (equation (3) ), sort of
> rescaled
> "truncated mean" ...
>
> 12) page 7
> "... has been studied in [12] using ..."
> I prefer "has been studied by xxx [12] using .." or simply
> ".. has been studied in reference [12] using .."
> But this is only matter of taste!
>
> 13) page 8 and 9
> The fig. 6 is very busy and all discussions regarding sigmas and mean
> values are difficult.
>
> 14) page 11
> " If we take into account that this estimate has a factor two
> uncertainty.."
> This phrase must be better justified.
>
> 15) page 11
> " ... the non linearity effects ... are below 20% level."
> This statement is in contradiction with " ... the synergetic use of the
> digital
> and analog information ... analysis cut below the onset of ERP
> non-linearity"
> at page 5 .
>
> 16) page 26
> An arrow indicating the Gamma-cut value could be added.
>
> 17) page 13
> "In the first period, three events are close ... three otherssurvive the
> cut"
> For these six events you should provide also the ERP energy, the beta
> (if available) and an event display with indications (if any) of delta
> rays.
>
> That's all.
>
> Ciao e buon lavoro
>
> Marco
This archive was generated by hypermail 2a24 : Mon Feb 05 2001 - 06:31:07 EST